I have a question regarding the generated dataset Further Education. According to the data manual, instances from spVocTrain which can be classified as courses and training related to further education are included in the generated dataset „Further Education“.
Hence the Further education dataset contains a subset of the respondents contained in the spVocTrain dataset. I wanted to know if there is a variable in the spVocTrain dataset that captures this information : whether the vocational training was a part of further education or was undertaken in the normal course of education.
there is no direct variable backtracking the source of episodes in FurtherEducation generated in spVocTrain. However, I think you can reconstruct such a variable yourself. Episodes were transferred if the type of the program is a „course to obtain a license“. The following syntax illustrates this:
use SC6_FurtherEducation.dta, clear
keep if tx28200==24 // rows from spVocTrain
tempfile fe
save `fe'
use SC6_spVocTrain.dta, clear
keep if subspell==0
keep if ts15291_g12==2 | ts15201==14 | ts15201_v1==13
merge 1:1 ID_t splink using `fe' // all cases match
Thank you. However I am trying to understand the reason behind this kind of filtering for the Further Education. According to the condtition you mentioned
I understand that people who undertook course for the license are included. However I dont understand why you included also ts15291_g12==2 ?
Also, I am trying to look at people who took a Voctrain course as a part of further training, which means courses undertaken after initial schooling regardless if they were to obtain a license or not. I use the following syntax
//Preparing spVocTrain
neps, language(en) : use "spVocTrain", clear
keep if subspell==0
fre ts15201 ts15201_v1
recode ts15201 (.m=.m "Missing by design") (.b=.b "Don't know") (.c=.c "Refused") (1=1 "Dual training") (2/8 12=2 "School-based (full-time) training") ///
(9=3 "University of applied sciences") (10/11 15/16=4 "University") ///
(13/14=5 "Further education") (17=6 "Another kind of training program"), gen(traintype)
recode ts15201_v1 (.m=.m "Missing by design") (.b=.b "Don't know") (.c=.c "Refused") (1=1 "Dual training") (2/5 11=2 "School-based (full-time) training") ///
(6=3 "University of applied sciences") (7/10=4 "University") ///
(12/14=5 "Further education"), gen(traintype2)
replace traintype = traintype2 if traintype2!=.m
lab var traintype "Type of post-school education"
keep if traintype==5 //only keeping instances of further education
save 'furthertraining'
Is this categorization for different type of training correct?
sorry, I haven’t been precise in my comment above:
Actually, FurtherEducation was constructed to contain all courses. Courses in spVocTrain are defined either if the type of the program is a „course to obtain a license“ (ts15201==14, ts15201_v1==13), or the vocational training was identified as a course during coding of the open-ended content variable (ts15291_g12==2). It then also has an entry in ts15291_g13 in most cases, so you can derive the kind of the course out of it.
I cannot help you with your second question, as this mainly depends on your setting/theory.